A lora i trained on a test set of images for a certain character/aesthetic which is mostly trained/focused on: - Black sclera - Colors with good contrast against them (red, blue, yellow, green, other bright colors). - Outfits that match well with that color scheme UPDATE: v3 <span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, &quot;system-ui&quot;, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">is a bit different than v2, more consistent but less variety, try it and use v2 if you don't like how it adheres. This one uses tacstyle0 as the trigger world, check the example images for a new example prompt. Dunno how it interacts with other loras and prompt types yet UPDATE: v2 can be activated to various extents with both the tacstyle and exampleuserstyle keywords Older ones only have tacstyle as an activation term.

Tacstyle - terms_and_conditions_style

Trained a new lora to 3000 steps but on the de-distilled version of the zturbo model. Prompt adherence is still pretty buggy but is significantly more stable. Can now be activated to various extents with both the tacstyle and exampleuserstyle keywords

Files with SHA256: e4f0dbbd00bd8bc85bb2285b482299d4f157603cfca82f63fcf4c7f650d1b28b